InstaGR8 - Instagram Viewer for MSX1 and MSX2 that have TCP-IP UNAPI Network Adapters

By ducasp

Hero (613)

ducasp's picture

25-05-2020, 23:59

First of all, this wouldn't exist at all without @glufke, he has developed, created, tested and make available the server infrastructure behind this as well as the concept MSX Basic code that works for GR8Net. If you like this, do not forget to say thanks and be grateful to @glufke as well, the original InstaGR8 can be found in this topic:

Also, I'm curious about how faster Glufke's version is on GR8NET, if anyone can compare, I would think it would be faster because uses GR8NET buffer that receives the whole file at once and then move to VRAM, all that done really fast by GR8NET, while on UNAPI we need to handle the HTTP protocol on the MSX, as well as send small blocks to the VRAM (from 512 bytes up to 1024 bytes, depending on your adapter TCP-IP buffer), a good comparison would be to check the time to load as an example, #sonic, then hit right to go to the next image, and check how much time from when right was hit to when the image and its description are fully presented. My guess is that glufke's Basic program will be faster and a better option for those that have a GR8NET, just curious to know how much better. ;)

You can download it from my github, as well as get its source code:

"Description" wrote:

This is an Instagram Viewer based on INSTAGR8 MSX Basic software that is a client to Glufke's server. His server will translate instagram pictures into MSX1 Screen2 or MSX2 Screen 8 compatible images. Those images, along with a few characters of the image description are shown on screen. This MSX-DOS client of InstaGR8 server has the advantage of not needing GR8NET, working on ObsoNet, Denyonet, SM-X WiFi, GR8NET and any other adapter for MSX that is TCP-IP UNAPI Compliant.

As it doesn't rely on GR8NET specific API, it is not possible to show the picture as fast as original Glufke program does. This is because GR8NET simply store the whole image in its own buffer and then transfer directly to VRAM. Meanwhile, in this program case, it needs to handle all the HTTP protocol to receive files, and fill the VRAM as it receives every 512 bytes or 1024 bytes, depending on your adapter TCP-IP buffer. If we are talking more specifically about Obsonet, then it is well known that it is not capable of much more than a couple KB/s transfers, and a Screen 8 picture is 53KB large, that means it will take quite a bit to finish a single image. On SM-X, that has a fully accelerated network stack, it is pretty fast, taking a couple of seconds. Now, for MSX1 images, those have 12KB, so loading times are faster.

Hope everyone that have other network adapters than GR8NET enjoy it!

Again, thanks @glufke for your hard work! :)

Login or register to post comments

By glufke

Resident (37)

glufke's picture

26-05-2020, 03:58

I will try both and share my findings this week. Thanks for the port!

By ducasp

Hero (613)

ducasp's picture

26-05-2020, 22:19

By the way, UNAPI version of Instagr8 is available on MSX HUB now:


Or using HUBG, it is in the Internet category Wink